mfkiwl/robust_cvd

Robust Consistent Video Depth Estimation

[CVPR 2021] Robust Consistent Video Depth Estimation

This repository contains Python and C++ implementation of Robust Consistent Video Depth, as described in the paper

Johannes Kopf, Xuejian Rong, and Jia-Bin Huang. Robust Consistent Video Despth Estimation. CVPR 2021

Project | Paper | Video | Colab

We present an algorithm for estimating consistent dense depth maps and camera poses from a monocular video. We integrate a learning-based depth prior, in the form of a convolutional neural network trained for single-image depth estimation, with geometric optimization, to estimate a smooth camera trajectory as well as detailed and stable depth reconstruction.

Changelog

• [June 2021] Released the companion Colab notebook.
• [June 2021] Initial release of Robust CVD.

Installation

Please refer to the colab notebook for how to install the dependencies.

Running

Please refer to the colab notebook for how to run the cli tool for now.

Result Folder Structure

frames.txt              # meta data about number of frames, image resolution and timestamps for each frame
color_full/             # extracted frames in the original resolution
color_down/             # extracted frames in the resolution for disparity estimation 
color_down_png/      
color_flow/             # extracted frames in the resolution for flow estimation
flow_list.json          # indices of frame pairs to finetune the model with
flow/                   # optical flow 
mask/                   # mask of consistent flow estimation between frame pairs.
vis_flow/               # optical flow visualization. Green regions contain inconsistent flow. 
vis_flow_warped/        # visualzing flow accuracy by warping one frame to another using the estimated flow. e.g., frame_000000_000032_warped.png warps frame_000032 to frame_000000.
depth_${model_type}/    # initial disparity estimation using the original monocular depth model before test-time training
R_hierarchical2_${model_type}/ 
    flow_list_0.20.json                 # indices of frame pairs passing overlap ratio test of threshold 0.2. Same content as ../flow_list.json.
    videos/                             # video visualization of results 
    B0.1_R1.0_PL1-0_LR0.0004_BS4_Oadam/
        checkpoints/                    # checkpoint after each epoch
        depth/                          # final disparity map results after finishing test-time training
        eval/                           # intermediate losses and disparity maps after each epoch 
        tensorboard/                    # tensorboard log for the test-time training process

Citation

If you find our work useful in your research, please consider citing:

@inproceedings{kopf2021rcvd,
 title={Robust Consistent Video Depth Estimation},
 author={Kopf, Johannes and Rong, Xuejian and Huang, Jia-Bin},
 year={2021},
 booktitle=IEEE/CVF Conference on Computer Vision and Pattern Recognition
}

License

See the LICENSE for more details.

Issues & Help

For help or issues using Robust CVD, please submit a GitHub issue or a PR request.

Before you do this, make sure you have checked CODE_OF_CONDUCT, CONTRIBUTING, ISSUE_TEMPLATE, and PR_TEMPLATE.

Acknowledgements

Check our previous work on Consistent Video Depth Estimation.

We also thank the authors for releasing PyTorch, Ceres Solver, OpenCV, Eigen, MiDaS, RAFT, and detectron2.